• Northumbria University Featured PhD Programmes
  • University of Glasgow Featured PhD Programmes
  • Staffordshire University Featured PhD Programmes
  • University of Manchester Featured PhD Programmes
  • University of Southampton Featured PhD Programmes
  • Queen’s University Belfast Featured PhD Programmes
  • University of East Anglia Featured PhD Programmes
  • University of Stirling Featured PhD Programmes
University of Warwick Featured PhD Programmes
University of Huddersfield Featured PhD Programmes
Imperial College London Featured PhD Programmes
Engineering and Physical Sciences Research Council Featured PhD Programmes
University College London Featured PhD Programmes

Incremental Non-Rigid 3D Reconstruction from Monocular Video

This project is no longer listed in the FindAPhD
database and may not be available.

Click here to search the FindAPhD database
for PhD studentship opportunities
  • Full or part time
    Dr J Wu
    Dr Y Lai
  • Application Deadline
    No more applications being accepted
  • Funded PhD Project (Students Worldwide)
    Funded PhD Project (Students Worldwide)

Project Description

Non-rigid 3D reconstruction has attracted great interest over the last decades, but is typically restricted to using multiple or depth-sensing cameras, or dealing with specific shapes. Reconstruction from a single RGB-only camera (known as NRSfM: non-rigid structure from motion) is significantly harder. However, the ability to do it would open the door to easy, lightweight 3D capture, and enable reconstruction from existing videos.
Most NRSfM approaches are either sparse feature-based, which leads to very low resolution 3D models, or batch methods unsuitable for real-time applications. Incremental dense reconstruction has been developed with RGB-depth streams, but not with RGB-only videos. Moreover, existing NRSfM approaches have limited ability to preserve shape details. Shape from shading (SFS) methods are known for their ability to reconstruct fine shape details, but are largely unexplored in NRSfM.
This project aims to advance the research in NRSfM by 1) introducing an incremental sequential reconstruction pipeline; and 2) incorporating SFS to improve the reconstruction of fine details. To meet these aims, the following specific objectives are set:
• To review the research in NRSfM, as well as techniques in relevant areas, e.g. image/video processing, multi-view stereo, SFS. [6 months]
• To devise techniques for differentiating points subject to rigid/non-rigid movements from successive frames, and establish point correspondence. The aim is to obtain an incrementally refined rigid template and reconstruct non-rigid shapes with its aid. [14 months]
• To explore methods for illumination/albedo estimation, and shading extraction at each frame, with the aid of the template. [6 months]
• To explore approaches enforcing the shading and illumination as additional constraints on the non-rigid reconstruction, in order to preserve fine details. [10 months]
The main supervisor has years of research experience in SFS, and the co-supervisor has expertise in 3D modelling, which will ensure smooth progress and timely completion of the project.

Funding Notes

Cardiff University funding of full UK/EU tuition fees (£4,121 per annum in 2016/17) and doctoral stipend matching UK Research Council National Minimum (£14,296 per annum in 2016/17). A successful overseas applicant would be required to pay the difference between UK/EU tuition fees (as above) and international tuition fees (£18,250 per annum in 2016/17). Applicants for a studentship must have obtained, or be about to obtain, a 2.1 degree or higher or a masters degree in a relevant subject.
More information: http://www.cs.cf.ac.uk/degreeprogrammes/research/

Related Subjects

How good is research at Cardiff University in Computer Science and Informatics?

FTE Category A staff submitted: 13.73

Research output data provided by the Research Excellence Framework (REF)

Click here to see the results for all UK universities

Cookie Policy    X